Free AI Referral Traffic Tracking in Analytics

Free analytics can show you the visits that arrive from AI apps and identify themselves. That means sessions with an AI app in the referrer, or with a tag like utm_source=chatgpt.com in the landing URL.

That answers a narrow, useful question: which pages AI users click through to, and what those visitors do next. It does not tell you how often AI answers mention you, which answers cite you, or how many people saw your name and never clicked.

Treat it as one input. If you are deciding whether this is worth the effort at all, see whether AI search traffic is worth tracking for a small business.

These steps use Google Analytics 4 and cost nothing. The same logic works in any tool that records referrer and UTM data.

1. Check what already exists. Open Reports > Acquisition > Traffic acquisition and switch the primary dimension to Session source / medium. Our evidence base notes Google Analytics added a native AI assistant channel in May 2026. If your property shows it, compare it with step 3 rather than replacing it blindly.
2. Search for the ChatGPT tag. In the same report, filter Session source for chatgpt.com. ChatGPT appends utm_source=chatgpt.com to the links it sends, so these sessions are tagged even when the browser strips the referrer. Read more on how UTM parameters identify AI referrals.
3. Build a custom channel. Go to Admin > Data display > Channel groups > Create new channel group. Add a channel called AI referrals with the condition Session source matches regex:

chatgpt\.com|openai\.com|perplexity\.ai|claude\.ai|gemini\.google\.com|copilot\.microsoft\.com

4. Order it above Referral. GA4 assigns each session to the first channel that matches. If Referral sits higher, your AI sessions disappear into it.
5. Add landing pages. Open an exploration with the dimensions Session custom channel group and Landing page, plus your key event as a metric. This shows which pages AI visitors reach and whether they convert.
6. Check the raw logs if you have them. Server or CDN logs record the Referer header for every request. They catch visitors who block the analytics script.

Re-check the regex every quarter. AI apps add and rename domains, and the hostnames above are a starting list, not a complete one.

Read the AI channel as counted visits from identifiable AI sources. Three habits keep the numbers honest.

• Compare sources, not shares of a market. ChatGPT tags its links, and most other apps do not. A large ChatGPT slice partly reflects that tagging, so it measures instrumentation as much as usage.
• Watch landing pages. If AI visitors land on pages you never promoted, those pages are the ones engines are linking to. Check that the facts on them (prices, plan names, locations) are current.
• Judge quality, not just volume. Small AI visit counts with high conversion can matter more than a large referral source that bounces. Our guide on whether you can attribute leads and sales to AI search covers how far that attribution can go.

A rising AI channel is good news, but it is not proof that your visibility rose. The app may have changed how it tags links, or a tagged surface may have gained users.

Several kinds of AI-driven visits never show up as AI referrals.

Google AI Overviews and AI Mode

Google treats AI Overviews and AI Mode as part of Search. A click from an AI Overview supporting link arrives as Google organic traffic, with no distinct referral to separate it. The only first-party view is the generative AI report in Search Console, which shows impressions only, with no clicks or position. See how to view AI Overview impressions free in Search Console.

Untagged app traffic

Even ChatGPT's tagging is partial. Free-tier and mobile app clicks frequently arrive with no tag and no referrer, and analytics files them as direct. This is why ChatGPT traffic often shows up as direct.

Visits with no click at all

Most of AI's effect on a brand happens inside the answer. A user reads your name, then searches for you later or types your URL. Analytics records that as branded search or direct, not AI.

Crawler visits

AI crawlers fetch your pages but do not run JavaScript, according to Vercel's crawler study, so script-based analytics never records them. That is why AI crawler visits don't appear in Google Analytics. Crawler activity lives in server logs, and it measures access, not human interest.

Referral data starts after a click. It cannot tell you whether ChatGPT, Perplexity or Google's AI features name you for the questions your buyers ask, or whether they describe you correctly. That is why Google Analytics alone can't tell you how visible you are in AI answers.

Answer-level checks have limits of their own. ChatGPT rewrites a user's prompt into its own search queries, and retrieval fires at the model's discretion, so any prompt sample is a sample. Use referral data and answer checks together, and expect them to disagree at times.

For the traffic half, the custom channel above is the free tool. Set it up once and it runs on its own.

For the visibility half, the free checker on this site runs up to three buyer questions on AI engines and shows whether your brand is named, as a one-off snapshot. Pair the two: the checker tells you whether you appear, and the AI referrals channel tells you whether anyone clicks. For a deeper method, read the full guide to measuring AI referral traffic in analytics.

What does utm_source=chatgpt.com mean in my analytics? It means the visit came from a link inside ChatGPT. OpenAI adds that parameter to referral URLs automatically, so the session is labelled even if the browser drops the referrer. You don't need to add it yourself. Its absence doesn't prove a visit wasn't from ChatGPT, because some app and free-tier clicks still arrive untagged and land in direct.

Can I see Google AI Overview clicks in Google Analytics? Not separately. Clicks from AI Overview and AI Mode links arrive as normal Google organic sessions, with no distinct referral. Search Console's generative AI report shows impressions for those surfaces, but no clicks, click-through rate or position. You can see that you were shown, and you can see total organic visits, but you can't split the AI share out.

Is the GA4 AI channel accurate enough to report to leadership? It's accurate for what it counts: visits from identifiable AI sources. Report it as a floor and say so. Label it AI referrals we can identify, note that AI Overviews and some app traffic are excluded, and avoid calling it AI's share of traffic. Trends are more useful than absolute numbers, as long as tagging behaviour hasn't changed.

Do I need paid tools to track AI referral traffic? No. Google Analytics 4, server logs, or any privacy-focused analytics that records referrers can do it for free. Paid tools mostly add prebuilt regex lists and dashboards. None of them can see AI Overview clicks as a separate source or recover untagged app visits, because that data never reaches your site.

Why did my AI referral traffic suddenly jump or drop? Check tagging and routing before assuming visibility changed. An app may have started or stopped adding UTM tags, changed its referrer policy, or launched a new domain your regex misses. Also check whether a single landing page drove the change. If the shift is real, an answer check for your main buyer questions helps confirm it.